1. Field of the Invention
Embodiments of the present invention relate generally to devices for generating electromagnetic radiation. More particularly, embodiments of the present invention include vacuum compatible high frequency electromagnetic wave source components and methods of micro-fabricating devices including such vacuum compatible high frequency electromagnetic wave source components.
2. State of the Art
Conventional methods for producing microwave and millimeter wavelength electromagnetic radiation are well known in the art. Such conventional methods typically involve the use of electron tubes that rely on various forms of velocity modulation of an electron beam. Electron beam modulation may then be followed by some form of drift used to achieve electron density bunching. After bunching, the electron kinetic energy may be converted into microwave or millimeter waves. In klystrons, this may be achieved using two-cavity and multi-cavity configurations. Klystrons tend to be bandwidth limited, however. To achieve wider bandwidths, traveling wave velocity modulation is used. For example, in a conventional traveling wave tube amplifier (TWTA) electrons interact with the longitudinal electric component of a slow electromagnetic wave. The phase velocity of the electromagnetic wave is slowed down to match the electron velocity. The slow wave structure (SWS) of a traveling wave tube (TWT) provides continuous and cumulative interaction between the electron beam and the electromagnetic wave, thereby producing microwaves or millimeter waves over bandwidths of an octave or more.
Klystrons are among the oldest of the electron velocity modulated devices dating back to the 1930s. Smaller klystrons find a number of uses where relatively very narrow bandwidth is acceptable. A class of miniature klystron devices is known as “Klystrinos”. Klystrinos utilize manufacturing techniques employing a photosensitive material which requires X-rays from an instrument known as synchrotron radiation source to expose the photosensitive materials. Besides the scarcity of such synchrotrons, structures fabricated by this technique require extensive additional and final machining to have the proper physical and electrical characteristics.
The surface-finish of the Klystrino output cavity is critical to the efficiency of the device. Good surface finish is required to obtain maximum quality factor, Qo. Circuit efficiency ηckt=Qo/(Qo+Qo), increases with increasing Qo. Early RF designs of the 95 GHz Klystrino circuit determined that the optimum external Q(Qc) in the output cavity was on the order of 250. The best Qo obtainable with normal machining or EDM was around 800 at 95 GHz. Circuit efficiency in this case would be 76%, implying that 316 Watts would be absorbed in the walls of the output cavity in order to realize a 1 kW average output, power. This would severely limit the average power achievable from the Klystrino. The theoretical maximum Qo for a copper cavity is around 1550. An alternative fabrication method was required to achieve an intrinsic Q close to the theoretical maximum.
LIGA fabrication has been considered the only process capable of producing copper cavities with surface finish good enough to approach the theoretical value for Q. LIGA is a German acronym for X-ray lithography, electrodeposition and molding. One drawback with LIGA fabrication is that it requires access to a synchrotron light source. Thus, a limited number of facilities exist that could produce LIGA processed substrates for a Klystrino.
Another drawback with the conventional LIGA fabrication process is exposing the polymethylmethacrylate (PMMA) photoresist to a depth of 1 mm. In order to produce 1 mm deep structures, a 25 micron thick gold X-ray mask is bonded to the top of the PMMA and the assembly is then exposed and etched repeatedly until the 1 mm depth is reached. The repeated exposure and etching steps and use of gold for an X-ray mask add to the cost of producing electronic devices using the LIGA process.
Another constraint sometimes encountered with LIGA processing is the need to use aluminum as the substrate for the LIGA process, as opposed to other substrate materials. Aluminum is desirable for LIGA processing because it has a low atomic number and, therefore, the backscattering of X-ray photons is minimized. Aluminum also has a high coefficient of expansion that tends to match the expansion of the PMMA photoresist. Additionally, the surface of an aluminum substrate can easily be roughened to provide better adhesion for the PMMA photoresist. The combination of roughened surface and high coefficient of expansion enables the bond between the aluminum and the PMMA photoresist to survive the repeated thermal cycling that occurs with multiple exposures. Backscattering can be a problem because photons backscattered from the substrate can expose the PMMA photoresist near the edge of the mask, resulting in PMMA columns that are undercut. If the backscattering is severe, the PMMA column will detach from the base during etching. Less severe backscattering will result in cavities with smaller volumes and correspondingly higher resonant frequencies.
Once the PMMA photoresist has been exposed and etched, the structure is copper plated until the un-etched PMMA is completely covered by the electrodeposited copper. As a result of the above constraints, neither the top nor the bottom surface of the electroplated LIGA part is suitable as the bottom wall of the Klystrino cavity. The bottom surface is roughened aluminum and the top surface still has the gold mask that was on top of the unexposed PMMA.
The number and complexity of the post-LIGA circuit fabrication steps are significantly increased due to the above issues. First, the top surface of the electrodeposited part must be machined (i.e., diamond flycut) to produce a flat reference plane for subsequent machining. Next, the aluminum is etched away in NaOH leaving a rough copper surface. The base of the cavities starts as a machined 1 mm thick copper sheet. Since brazing would leave fillets at the edges of the cavities that would modify the cavity frequencies, diffusion bonding is used to fuse the LIGA structure to the copper base. It is important that there are no unbonded areas at the edges of the cavities as that would lower the Q dramatically.
Once bonding is complete, the slots for the iron polepieces are cut using, for example, wire electrical discharge machining (EDM). It is important that the polepiece slots be aligned with the centerline of the cavities. The circuit with cavities and polepiece slots is then brazed to the cooling and support structure. Since this part houses the magnets, it also requires very accurate alignment with the RF circuit.
With all the brazing steps complete, the assembly is now ready for final machining. Final machining consists of three parts. First the LIGA section is machined to a height of 1 mm. The tuning rate for this dimension is 30 MHz/micron, so an error of 0.0002″ in this operation will shift the cavity resonant frequencies by 150 MHz. The second step is the milling of the beam tunnel. A ball endmill can be used to cut the 800 micron diameter beam tunnel into each half of the Klystrino circuit. The final machining operation cuts the coupling irises for the input and output waveguides, the coupling slots for the five-gap output cavity, the input and output waveguides and the vacuum pumpout channels to eliminate gas pockets. Measurements of intrinsic Q's for the LIGA fabricated cavities ranged from 1300 to 1500.
At this point in the fabrication process, the cavity frequencies can be measured in cold test to determine if any frequency adjustment is necessary. For example, in the original Klystrino it was determined that several cavities needed to be tuned. This was done using a 0.010″ end mill in a high speed spindle on a CNC mill. One half of the circuit was machined to produce half the desired change in resonant frequency. The parts were cold tested again and the volume of material to be removed in the opposite circuit half was adjusted based on the cold test results. Cavities which resonated at a lower frequency than desired were adjusted by increasing the width of the gap. Cavities with higher than desired frequencies were adjusted by cutting a racetrack slot in the back wall of the cavity to increase cavity volume.
Once the resonant frequencies were achieved, the magnets and polepieces were inserted into the two cavity halves. The circuit halves were bolted together and the sides and ends of the cavity block were machined to accept the waveguides and gun polepiece. Additional parts such as the electron gun and collector are then attached. The completed Klystrino is installed in a vacuum vessel that is evacuated or into a vacuum package that is evacuated and sealed. This technique also applies to other types of devices including, but not limited to, traveling wave tubes (TWTs), back wave oscillators (BWOs), magnetrons, klystrons, and other millimeter wave and microwave devices.
Electromagnetic radiation sources at millimeter wavelengths encounter significant problems during manufacturing for two reasons. First, the device dimensions vary inversely with operating frequency. Second, as the frequency increases, skin depths shrink and circuit losses increase. This means that the surface finish of the electromagnetic circuits must be improved while the fabrication tolerances become more difficult to achieve. These problems suggest the need for alternative circuit fabrication methods that are significantly different from conventional lathe and mill machining used in lower frequency devices.
Thus, there still exists a need in the art for improved vacuum compatible high frequency electromagnetic and millimeter wave sources, methods of micro-fabricating devices including vacuum compatible high frequency electromagnetic wave sources.